I'm trying to do a positive clamper circuit in easyeda.com to bias a triangular signal of 200mV pk-to-pk at 1kHz centered around 0V. Here's the circuit:

Clamper circuit

The bias seems to work only during the first cycle:

Slow decay of bias during the first cycles.

But as you can see, the bias exponentially decreases slowly and after ~1s, it's almost at the same DC as the original signal. What I want is a circuit that bias the signal for half of the pk-to-pk so that it always stays like the orange signal in the 2nd picture.

Exponential decay of the bias.

The triangular signal's frequency could be from 1Hz up to 1kHz. Why does the bias decreases in time? Would you suggest another circuit to bias this signal without the decay that would work for the frequency range I want? Does it has to do with the values of R and C?

  • \$\begingroup\$ Check if the diode is getting forward biased. 200mV seems too small to forward bias the diode during negative cycle \$\endgroup\$ – AJN Jul 15 '20 at 3:23
  • \$\begingroup\$ Welcome. To move your signal positive you need to add voltage through a 10 K or higher resistor. Your voltage source is the offset you want. The diode only acts as a negative voltage clamp. \$\endgroup\$ – user105652 Jul 15 '20 at 3:23
  • \$\begingroup\$ The reason why it seems to work for first few cycles may be that the capacitor has initial voltage set on it (either manually or automatically by the simulator). \$\endgroup\$ – AJN Jul 15 '20 at 3:33

During the negative cycles, the capacitor needs to charge so that a positive potential develops on its right side plate. This voltage will add to the waveform during rest of the operation of the circuit.

To charge the capacitor during negative half cycles only, the diode needs to be conducting during negative half cycles. -100mV amplitude is not enough to forward bias the diode and make it conducting.

Repeat your experiment with say 2V amplitude triangle wave and see if the situation improves.

To work with low voltages, you need either

  1. A precision clamper circuit
  2. A diode with a much smaller forward voltage drop.
  3. A voltage source which biases the diode to the threshold of conduction (or larger value ). Indicated as \$V_{ref}\$ in the linked Wikipedia page.


simulate this circuit – Schematic created using CircuitLab

  • \$\begingroup\$ Thank you for your answer, from what I saw in my simulations, I think it's the bias of the diode that caused my problem as my signal is ~100-200mV of amplitude. I was having trouble biaising the diode with a voltage source and I think it also depends on the amplitude of my signal. However, the precision clamper circuit with the op amp on wikipedia was really and it did the trick! It does exactly what I want. Thank you! \$\endgroup\$ – Jérémy Talbot-Pâquet Jul 15 '20 at 15:44

With an input signal amplitude of 200 mV p-p, the diode is ineffective at clamping because it just won't conduct any significant amount at these voltages. Here's a picture from another post that shows how much current a 1N4148 can be expected to take at various low voltages: -

enter image description here

The Y axis is the diode forward voltage and the X axis is the current it takes. With only 0.1 volts applied (the peak signal), the diode conducts 40 nA and is therefore totally ineffective at these low voltages as a clamp.

  • \$\begingroup\$ Nice graph..... \$\endgroup\$ – Tony Stewart EE75 Jul 15 '20 at 12:55
  • \$\begingroup\$ So if the minimum diode impedance is MΩ , how would U explain his discrepancy between schematic and scope with output jump to 0 to 200mV? \$\endgroup\$ – Tony Stewart EE75 Jul 15 '20 at 13:13
  • \$\begingroup\$ @TonyStewartSunnyskyguyEE75 I don't need to explain it; you have done this already. \$\endgroup\$ – Andy aka Jul 15 '20 at 13:16

There are a couple false assumptions:

  1. Sig Gen is not centred around 0V but rather 100mVpk + 100mVdc so 0 to 200mV. The output is identical starting off as you should expect.
  2. The 100mV peak steady state on the diode conducts only about 1uA momentarily then almost nothing the rest of the time so the 250k slowly biases the output to 0V. If that were 10M, you might see some DC bias.

Here is an "active clamp" circuit using a comparator function driving a transistor active clamp then buffered to makes a more ideal clamp circuit that even works on 10mV signals.

  • \$\begingroup\$ regarding point #1, the blue line seems to be the input and is varying b/w -100mV and 100mV. There doesn't seem to be any DC. \$\endgroup\$ – AJN Jul 15 '20 at 6:07
  • \$\begingroup\$ Not possible with circuit shown and I/O waveforms \$\endgroup\$ – Tony Stewart EE75 Jul 15 '20 at 13:11
  • \$\begingroup\$ Hi Tony, thanks for your answer, I tried in easyeda, but the active clamp with transistor didn't work so I couldn't see what it is supposed to do. What would be the advantage of using this over the precision clamper circuit shown on wikipedia? en.wikipedia.org/wiki/Clamper_(electronics)#Op-amp_circuit \$\endgroup\$ – Jérémy Talbot-Pâquet Jul 15 '20 at 15:39
  • \$\begingroup\$ Didn't my link simulate ok in your browser? easyEDA must be buggy \$\endgroup\$ – Tony Stewart EE75 Jul 15 '20 at 17:41
  • \$\begingroup\$ The WIki Clamp shown has high impedance =R and diode orientation shown is a +ve clamp with negative output but also needs a negative supply voltage. Mine is a buffered single supply using CMOS OA \$\endgroup\$ – Tony Stewart EE75 Jul 15 '20 at 17:50

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.